In
<42B56C6A683EBA4581C01CB49A914CA702916273(_at_)MAILFAXSRV(_dot_)gfimalta(_dot_)com>
"Stefan Engelbert" <stefan(_at_)gfi(_dot_)com> writes:
I dont like PRA at all - but you cannot say its untested.
Considering hat PRA is more or less the same as the fetchmail
algorithm...
No, the PRA has important differences from the fetchmail algorithm.
In particular, fetchmail checks for the Return-Path: header, and the
rest of it is a fallback in case you actually get a malformed email
that doesn't have one.
The fetchmail algorthim has never been used to pass stuff to the the
SPF check_host() function.
I can say that the PRA is untested because I've long looked for data
on how it performs and I can only find very limited testing.
(E.g. the stuff by Andy Newton and Mark Lentczner where they checked
their personal mailboxes, and more recently, a little feedback from
the sendmail SID-milter folks.)
-wayne